KSU paired amplicon community ordinations

Comparative NMDS plots of 16S and ITS data (EDGE samples excluded), updated to include adonis2 analyses (permANOVA by metadata)

Back to Home

opts <- options(knitr.kable.NA = "")

Set-up

0.1 Import clean OTU tables

0.2 Ordinate all grasses and by-host (trying k 1-7)

0.3 QC, stress check and validation of ordinations

ord_checkfit_df = data.frame()

for (name in comm.ord.names.by.grass) {
  
  output = extract.nmds.checkfit(get(name))

  ord_checkfit_df = rbind(ord_checkfit_df, output)

}

# Ideal stress value is < 0.05 (but k > 5 makes interpretation difficult)
    # Higher than 0.2 is poor (risks for false interpretation).
    # 0.1 - 0.2 is fair (some distances can be misleading for interpretation).
    # 0.05 - 0.1 is good (can be confident in inferences from plot).
    # Less than 0.05 is excellent (this can be rare).


ord_checkfit_df$stress<-round(ord_checkfit_df$stress,3)
ord_checkfit_df$nonmetric.fit.r2<-round(ord_checkfit_df$nonmetric.fit.r2,3)
ord_checkfit_df$linear.fit.r2<-round(ord_checkfit_df$linear.fit.r2,3)


# Final choices on k values

## AllG
  subset(ord_checkfit_df,ord_checkfit_df$input_data==comm.names.by.grass[1])
  subset(ord_checkfit_df,ord_checkfit_df$input_data==comm.names.by.grass[2])
  
  # flipping x axis coords for ease of interpretation
    # Fun.AllG.k5.ord$points[,1]<-Fun.AllG.k5.ord$points[,1]*-1 # already done
  
  saveRDS(Bac.AllG.k5.ord, file="../processed_data/clean_rds_saves/Bac.AllG.k5.ord.rds")
  saveRDS(Fun.AllG.k5.ord, file="../processed_data/clean_rds_saves/Fun.AllG.k5.ord.rds")

## ANGE
  subset(ord_checkfit_df,ord_checkfit_df$input_data==comm.names.by.grass[3])
  subset(ord_checkfit_df,ord_checkfit_df$input_data==comm.names.by.grass[4])
  
  #flip x coord of ITS Fun.ANGE.k5.ord
     # Fun.ANGE.k5.ord$points[,1]<-Fun.ANGE.k5.ord$points[,1]*-1 # already done
  
  saveRDS(Bac.ANGE.k5.ord, file="../processed_data/clean_rds_saves/Bac.ANGE.k5.ord.rds")
  saveRDS(Fun.ANGE.k5.ord, file="../processed_data/clean_rds_saves/Fun.ANGE.k5.ord.rds")

## BOER
  saveRDS(Bac.BOER.k5.ord, file="../processed_data/clean_rds_saves/Bac.BOER.k5.ord.rds")
  saveRDS(Fun.BOER.k5.ord, file="../processed_data/clean_rds_saves/Fun.BOER.k5.ord.rds")

## BOGR
  subset(ord_checkfit_df,ord_checkfit_df$input_data==comm.names.by.grass[7])
  subset(ord_checkfit_df,ord_checkfit_df$input_data==comm.names.by.grass[8])
  
  saveRDS(Bac.BOGR.k5.ord, file="../processed_data/clean_rds_saves/Bac.BOGR.k5.ord.rds")
  saveRDS(Fun.BOGR.k5.ord, file="../processed_data/clean_rds_saves/Fun.BOGR.k5.ord.rds")

## BUDA
  subset(ord_checkfit_df,ord_checkfit_df$input_data==comm.names.by.grass[9])
  subset(ord_checkfit_df,ord_checkfit_df$input_data==comm.names.by.grass[10])
  
  # flipping y axis coords for ease of interpretation
    # Fun.BUDA.k6.ord$points[,2]<-Fun.BUDA.k6.ord$points[,2]*-1 # already done
  
  saveRDS(Bac.BUDA.k6.ord, file="../processed_data/clean_rds_saves/Bac.BUDA.k6.ord.rds")
  saveRDS(Fun.BUDA.k6.ord, file="../processed_data/clean_rds_saves/Fun.BUDA.k6.ord.rds")

## SCSC
  subset(ord_checkfit_df,ord_checkfit_df$input_data==comm.names.by.grass[11])
  subset(ord_checkfit_df,ord_checkfit_df$input_data==comm.names.by.grass[12])
  
  saveRDS(Bac.SCSC.k5.ord, file="../processed_data/clean_rds_saves/Bac.SCSC.k5.ord.rds")
  saveRDS(Fun.SCSC.k5.ord, file="../processed_data/clean_rds_saves/Fun.SCSC.k5.ord.rds")

0.4 Set-up metadata

Available metadata:

 [1] "SampleID"       "ITS_ID"         "sample_name"   
 [4] "Replicate"      "Site"           "Site_Grass"    
 [7] "Site.Rep"       "Grass"          "Grass_Rep"     
[10] "Edge"           "SiteName"       "MAT_degC"      
[13] "MAP_mm"         "CollectionDate" "Latitude"      
[16] "Longitude"      "Elevation_m"    "Grassland"     
[19] "Gradient"       "Bin"            "julianday"     
[22] "coll_GDD"       "GDD30yr"        "GDD3yr"        
[25] "GDD2015"        "ppt30yr"        "ppt3yr"        
[28] "ppt2015"        "soil_moisture"  "GWC"           
[31] "perc_moisture"  "pH"             "phos"          
[34] "ammonium"       "nitrate"        "SOM"           
[37] "herbivory_perc" "avg_SLA"        "avg_SRL"       

SampleID

ITS_ID

sample_name

Replicate

Site

Site_Grass

Site.Rep

Grass

Grass_Rep

Edge

SiteName

MAT_degC

MAP_mm

CollectionDate

Latitude

Longitude

Elevation_m

Grassland

Gradient

Bin

julianday

coll_GDD

GDD30yr

GDD3yr

GDD2015

ppt30yr

ppt3yr

ppt2015

soil_moisture

GWC

perc_moisture

pH

phos

ammonium

nitrate

SOM

herbivory_perc

avg_SLA

avg_SRL

adonis2 tables

by grass

by Bin

by Gradient

save adonis2 outputs

Generate plots for later viewing

color by grass

color by bin

color by gradient

1 Data overview

All EDGE samples excluded
OTU tables rarefied (16S: 9,048 & ITS: 10,116 seqs/sample)
‘Taxa’ are OTUs

AllG (All Grasses)

Bacterial & Archaeal Dataset:

Bac.AllG

phyloseq-class experiment-level object
otu_table()   OTU Table:         [ 23156 taxa and 533 samples ]
sample_data() Sample Data:       [ 533 samples by 39 sample variables ]
tax_table()   Taxonomy Table:    [ 23156 taxa by 6 taxonomic ranks ]

Fungal Dataset:

Fun.AllG

phyloseq-class experiment-level object
otu_table()   OTU Table:         [ 6749 taxa and 509 samples ]
sample_data() Sample Data:       [ 509 samples by 39 sample variables ]
tax_table()   Taxonomy Table:    [ 6749 taxa by 7 taxonomic ranks ]

ANGE (Andropogon gerardii / Big bluestem)

Bacterial & Archaeal Dataset:

Bac.ANGE

phyloseq-class experiment-level object
otu_table()   OTU Table:         [ 14268 taxa and 106 samples ]
sample_data() Sample Data:       [ 106 samples by 39 sample variables ]
tax_table()   Taxonomy Table:    [ 14268 taxa by 6 taxonomic ranks ]

Fungal Dataset:

Fun.ANGE

phyloseq-class experiment-level object
otu_table()   OTU Table:         [ 2985 taxa and 98 samples ]
sample_data() Sample Data:       [ 98 samples by 39 sample variables ]
tax_table()   Taxonomy Table:    [ 2985 taxa by 7 taxonomic ranks ]

BOER (B. eriopoda / Black grama)

Bacterial & Archaeal Dataset:

Bac.BOER

phyloseq-class experiment-level object
otu_table()   OTU Table:         [ 8374 taxa and 84 samples ]
sample_data() Sample Data:       [ 84 samples by 39 sample variables ]
tax_table()   Taxonomy Table:    [ 8374 taxa by 6 taxonomic ranks ]

Fungal Dataset:

Fun.BOER

phyloseq-class experiment-level object
otu_table()   OTU Table:         [ 1662 taxa and 78 samples ]
sample_data() Sample Data:       [ 78 samples by 39 sample variables ]
tax_table()   Taxonomy Table:    [ 1662 taxa by 7 taxonomic ranks ]

BOGR (B. gracilis / Blue grama)

Bacterial & Archaeal Dataset:

Bac.BOGR

phyloseq-class experiment-level object
otu_table()   OTU Table:         [ 10108 taxa and 86 samples ]
sample_data() Sample Data:       [ 86 samples by 39 sample variables ]
tax_table()   Taxonomy Table:    [ 10108 taxa by 6 taxonomic ranks ]

Fungal Dataset:

Fun.BOGR

phyloseq-class experiment-level object
otu_table()   OTU Table:         [ 2668 taxa and 92 samples ]
sample_data() Sample Data:       [ 92 samples by 39 sample variables ]
tax_table()   Taxonomy Table:    [ 2668 taxa by 7 taxonomic ranks ]

BUDA (Bouteloua dactyloides / Buffalograss)

Bacterial & Archaeal Dataset:

Bac.BUDA

phyloseq-class experiment-level object
otu_table()   OTU Table:         [ 11048 taxa and 83 samples ]
sample_data() Sample Data:       [ 83 samples by 39 sample variables ]
tax_table()   Taxonomy Table:    [ 11048 taxa by 6 taxonomic ranks ]

Fungal Dataset:

Fun.BUDA

phyloseq-class experiment-level object
otu_table()   OTU Table:         [ 2369 taxa and 80 samples ]
sample_data() Sample Data:       [ 80 samples by 39 sample variables ]
tax_table()   Taxonomy Table:    [ 2369 taxa by 7 taxonomic ranks ]

SCSC (Schizachyrium scoparium / Little bluestem)

Bacterial & Archaeal Dataset:

Bac.SCSC

phyloseq-class experiment-level object
otu_table()   OTU Table:         [ 17568 taxa and 174 samples ]
sample_data() Sample Data:       [ 174 samples by 39 sample variables ]
tax_table()   Taxonomy Table:    [ 17568 taxa by 6 taxonomic ranks ]

Fungal Dataset:

Fun.SCSC

phyloseq-class experiment-level object
otu_table()   OTU Table:         [ 4347 taxa and 161 samples ]
sample_data() Sample Data:       [ 161 samples by 39 sample variables ]
tax_table()   Taxonomy Table:    [ 4347 taxa by 7 taxonomic ranks ]

2 NMDS By Grass Host

- ANGE (Andropogon gerardii / Big bluestem)
- SCSC (Schizachyrium scoparium / Little bluestem)
- BOER (B. eriopoda / Black grama)
- BOGR (B. gracilis / Blue grama)
- BUDA (Bouteloua dactyloides / Buffalograss)

NMDS plots of all samples, colored by grass host

knitr::kable((as.data.frame(subset(adonis.out.array_predGrass, 
                    comm==comm.names.by.grass[1])))[,1:5], "simple", 
             digits =c (0,1,3,2,3), 
             align = 'c')

	Df	SumOfSqs	R2	F	Pr(>F)
Bac.AllG.Grass	4	18.6	0.132	20.12	0.001
Bac.AllG.Residual	528	122.2	0.868
Bac.AllG.Total	532	140.8	1.000

knitr::kable((as.data.frame(subset(adonis.out.array_predGrass, 
                    comm==comm.names.by.grass[2])))[,1:5], "simple", 
             digits =c (0,1,3,2,3), 
             align = 'c')

	Df	SumOfSqs	R2	F	Pr(>F)
Fun.AllG.Grass	4	13.2	0.064	8.67	0.001
Fun.AllG.Residual	504	192.5	0.936
Fun.AllG.Total	508	205.7	1.000

3 Subset by Grass Host

3.1 Influence of discrete latitudinal bins

AllG (All Grasses)

NMDS plots of all samples, colored by latitudinal bin

knitr::kable((as.data.frame(subset(adonis.out.array_predBin, 
                    comm==comm.names.by.grass[1])))[,1:5], "simple", 
             digits =c (0,1,3,2,3), 
             align = 'c')

	Df	SumOfSqs	R2	F	Pr(>F)
Bac.AllG.Bin	3	7.9	0.056	10.48	0.001
Bac.AllG.Residual	529	132.9	0.944
Bac.AllG.Total	532	140.8	1.000

knitr::kable((as.data.frame(subset(adonis.out.array_predBin, 
                    comm==comm.names.by.grass[2])))[,1:5], "simple", 
             digits =c (0,1,3,2,3), 
             align = 'c')

	Df	SumOfSqs	R2	F	Pr(>F)
Fun.AllG.Bin	3	11.3	0.055	9.79	0.001
Fun.AllG.Residual	505	194.4	0.945
Fun.AllG.Total	508	205.7	1.000

ANGE (Andropogon gerardii / Big bluestem)

NMDS plots of ANGE host-associated samples, colored by latitudinal bin

knitr::kable((as.data.frame(subset(adonis.out.array_predBin, 
                    comm==comm.names.by.grass[3])))[,1:5], "simple", 
             digits =c (0,1,3,2,3), 
             align = 'c')

	Df	SumOfSqs	R2	F	Pr(>F)
Bac.ANGE.Bin	3	4.4	0.157	6.33	0.001
Bac.ANGE.Residual	102	23.8	0.843
Bac.ANGE.Total	105	28.2	1.000

knitr::kable((as.data.frame(subset(adonis.out.array_predBin, 
                    comm==comm.names.by.grass[4])))[,1:5], "simple", 
             digits =c (0,1,3,2,3), 
             align = 'c')

	Df	SumOfSqs	R2	F	Pr(>F)
Fun.ANGE.Bin	3	6.4	0.174	6.59	0.001
Fun.ANGE.Residual	94	30.6	0.826
Fun.ANGE.Total	97	37.0	1.000

BOER (B. eriopoda / Black grama)

NMDS plots of BOER host-associated samples, colored by latitudinal bin

knitr::kable((as.data.frame(subset(adonis.out.array_predBin, 
                    comm==comm.names.by.grass[5])))[,1:5], "simple", 
             digits =c (0,1,3,2,3), 
             align = 'c')

	Df	SumOfSqs	R2	F	Pr(>F)
Bac.BOER.Bin	3	6	0.354	14.58	0.001
Bac.BOER.Residual	80	11	0.646
Bac.BOER.Total	83	17	1.000

knitr::kable((as.data.frame(subset(adonis.out.array_predBin, 
                    comm==comm.names.by.grass[6])))[,1:5], "simple", 
             digits =c (0,1,3,2,3), 
             align = 'c')

	Df	SumOfSqs	R2	F	Pr(>F)
Fun.BOER.Bin	3	5.4	0.194	5.94	0.001
Fun.BOER.Residual	74	22.3	0.806
Fun.BOER.Total	77	27.7	1.000

BOGR (B. gracilis / Blue grama)

NMDS plots of BOGR host-associated samples, colored by latitudinal bin

knitr::kable((as.data.frame(subset(adonis.out.array_predBin, 
                    comm==comm.names.by.grass[7])))[,1:5], "simple", 
             digits =c (0,1,3,2,3), 
             align = 'c')

	Df	SumOfSqs	R2	F	Pr(>F)
Bac.BOGR.Bin	3	2.9	0.168	5.51	0.001
Bac.BOGR.Residual	82	14.3	0.832
Bac.BOGR.Total	85	17.2	1.000

knitr::kable((as.data.frame(subset(adonis.out.array_predBin, 
                    comm==comm.names.by.grass[8])))[,1:5], "simple", 
             digits =c (0,1,3,2,3), 
             align = 'c')

	Df	SumOfSqs	R2	F	Pr(>F)
Fun.BOGR.Bin	3	4.0	0.126	4.24	0.001
Fun.BOGR.Residual	88	27.7	0.874
Fun.BOGR.Total	91	31.8	1.000

BUDA (Bouteloua dactyloides / Buffalograss)

NMDS plots of BUDA host-associated samples, colored by latitudinal bin

knitr::kable((as.data.frame(subset(adonis.out.array_predBin, 
                    comm==comm.names.by.grass[9])))[,1:5], "simple", 
             digits =c (0,1,3,2,3), 
             align = 'c')

	Df	SumOfSqs	R2	F	Pr(>F)
Bac.BUDA.Bin	3	2.6	0.171	5.45	0.001
Bac.BUDA.Residual	79	12.4	0.829
Bac.BUDA.Total	82	15.0	1.000

knitr::kable((as.data.frame(subset(adonis.out.array_predBin, 
                    comm==comm.names.by.grass[10])))[,1:5], "simple", 
             digits =c (0,1,3,2,3), 
             align = 'c')

	Df	SumOfSqs	R2	F	Pr(>F)
Fun.BUDA.Bin	3	4.3	0.14	4.11	0.001
Fun.BUDA.Residual	76	26.5	0.86
Fun.BUDA.Total	79	30.8	1.00

SCSC (Schizachyrium scoparium / Little bluestem)

NMDS plots of SCSC host-associated samples, colored by latitudinal bin

knitr::kable((as.data.frame(subset(adonis.out.array_predBin, 
                    comm==comm.names.by.grass[11])))[,1:5], "simple", 
             digits =c (0,1,3,2,3), 
             align = 'c')

	Df	SumOfSqs	R2	F	Pr(>F)
Bac.SCSC.Bin	3	4.0	0.09	5.58	0.001
Bac.SCSC.Residual	170	40.8	0.91
Bac.SCSC.Total	173	44.8	1.00

knitr::kable((as.data.frame(subset(adonis.out.array_predBin, 
                    comm==comm.names.by.grass[12])))[,1:5], "simple", 
             digits =c (0,1,3,2,3), 
             align = 'c')

	Df	SumOfSqs	R2	F	Pr(>F)
Fun.SCSC.Bin	3	5.3	0.081	4.62	0.001
Fun.SCSC.Residual	157	60.0	0.919
Fun.SCSC.Total	160	65.3	1.000

3.2 Influence of discrete longitudinal gradients

AllG (All Grasses)

NMDS plots of all samples, colored by longitudinal gradient

knitr::kable((as.data.frame(subset(adonis.out.array_predGradient, 
                    comm==comm.names.by.grass[1])))[,1:5], "simple", 
             digits =c (0,1,3,2,3), 
             align = 'c')

	Df	SumOfSqs	R2	F	Pr(>F)
Bac.AllG.Gradient	2	17.5	0.124	37.57	0.001
Bac.AllG.Residual	530	123.3	0.876
Bac.AllG.Total	532	140.8	1.000

knitr::kable((as.data.frame(subset(adonis.out.array_predGradient, 
                    comm==comm.names.by.grass[2])))[,1:5], "simple", 
             digits =c (0,1,3,2,3), 
             align = 'c')

	Df	SumOfSqs	R2	F	Pr(>F)
Fun.AllG.Gradient	2	10.4	0.05	13.45	0.001
Fun.AllG.Residual	506	195.3	0.95
Fun.AllG.Total	508	205.7	1.00

ANGE (Andropogon gerardii / Big bluestem)

NMDS plots of ANGE host-associated samples, colored by longitudinal gradient

knitr::kable((as.data.frame(subset(adonis.out.array_predGradient, 
                    comm==comm.names.by.grass[3])))[,1:5], "simple", 
             digits =c (0,1,3,2,3), 
             align = 'c')

	Df	SumOfSqs	R2	F	Pr(>F)
Bac.ANGE.Gradient	1	2.8	0.1	11.62	0.001
Bac.ANGE.Residual	104	25.4	0.9
Bac.ANGE.Total	105	28.2	1.0

knitr::kable((as.data.frame(subset(adonis.out.array_predGradient, 
                    comm==comm.names.by.grass[4])))[,1:5], "simple", 
             digits =c (0,1,3,2,3), 
             align = 'c')

	Df	SumOfSqs	R2	F	Pr(>F)
Fun.ANGE.Gradient	1	1.8	0.05	5.04	0.001
Fun.ANGE.Residual	96	35.2	0.95
Fun.ANGE.Total	97	37.0	1.00

BOER (B. eriopoda / Black grama)

[NA on stats here - only 1 factor level]

# knitr::kable((as.data.frame(subset(adonis.out.array_predGradient, 
#                     comm==comm.names.by.grass[5])))[,1:5], "simple", 
#              digits =c (0,1,3,2,3), 
#              align = 'c')

# knitr::kable((as.data.frame(subset(adonis.out.array_predGradient, 
#                     comm==comm.names.by.grass[6])))[,1:5], "simple", 
#              digits =c (0,1,3,2,3), 
#              align = 'c')

BOGR (B. gracilis / Blue grama)

NMDS plots of BOGR host-associated samples, colored by longitudinal gradient

knitr::kable((as.data.frame(subset(adonis.out.array_predGradient, 
                    comm==comm.names.by.grass[7])))[,1:5], "simple", 
             digits =c (0,1,3,2,3), 
             align = 'c')

	Df	SumOfSqs	R2	F	Pr(>F)
Bac.BOGR.Gradient	2	2.6	0.152	7.45	0.001
Bac.BOGR.Residual	83	14.5	0.848
Bac.BOGR.Total	85	17.2	1.000

knitr::kable((as.data.frame(subset(adonis.out.array_predGradient, 
                    comm==comm.names.by.grass[8])))[,1:5], "simple", 
             digits =c (0,1,3,2,3), 
             align = 'c')

	Df	SumOfSqs	R2	F	Pr(>F)
Fun.BOGR.Gradient	2	3.5	0.109	5.46	0.001
Fun.BOGR.Residual	89	28.3	0.891
Fun.BOGR.Total	91	31.8	1.000

BUDA (Bouteloua dactyloides / Buffalograss)

NMDS plots of BUDA host-associated samples, colored by longitudinal gradient

knitr::kable((as.data.frame(subset(adonis.out.array_predGradient, 
                    comm==comm.names.by.grass[9])))[,1:5], "simple", 
             digits =c (0,1,3,2,3), 
             align = 'c')

	Df	SumOfSqs	R2	F	Pr(>F)
Bac.BUDA.Gradient	2	2.6	0.172	8.3	0.001
Bac.BUDA.Residual	80	12.4	0.828
Bac.BUDA.Total	82	15.0	1.000

knitr::kable((as.data.frame(subset(adonis.out.array_predGradient, 
                    comm==comm.names.by.grass[10])))[,1:5], "simple", 
             digits =c (0,1,3,2,3), 
             align = 'c')

	Df	SumOfSqs	R2	F	Pr(>F)
Fun.BUDA.Gradient	2	3.0	0.097	4.14	0.001
Fun.BUDA.Residual	77	27.8	0.903
Fun.BUDA.Total	79	30.8	1.000

SCSC (Schizachyrium scoparium / Little bluestem)

NMDS plots of SCSC host-associated samples, colored by longitudinal gradient

knitr::kable((as.data.frame(subset(adonis.out.array_predGradient, 
                    comm==comm.names.by.grass[11])))[,1:5], "simple", 
             digits =c (0,1,3,2,3), 
             align = 'c')

	Df	SumOfSqs	R2	F	Pr(>F)
Bac.SCSC.Gradient	2	5.9	0.131	12.93	0.001
Bac.SCSC.Residual	171	38.9	0.869
Bac.SCSC.Total	173	44.8	1.000

knitr::kable((as.data.frame(subset(adonis.out.array_predGradient, 
                    comm==comm.names.by.grass[12])))[,1:5], "simple", 
             digits =c (0,1,3,2,3), 
             align = 'c')

	Df	SumOfSqs	R2	F	Pr(>F)
Fun.SCSC.Gradient	2	3.4	0.052	4.37	0.001
Fun.SCSC.Residual	158	61.8	0.948
Fun.SCSC.Total	160	65.3	1.000